Recently, immuno-oncologists have turned their attention to the role of the second arm of the immune response—the more rough-and-ready innate arm, which serves as the body’s frontline defense against pathogenic invaders and, it seems, cancer.
Most anticancer immunotherapies that have reached clinical practice have focused on exploiting T cells, the major effectors of the adaptive immune response, with the goal of provoking an immune memory that leads to dramatic, long-lasting effects.
More recently, immuno-oncologists have turned their attention to the role of the second arm of the immune response—the more rough-and-ready innate arm, which serves as the body’s frontline defense against pathogenic invaders and, it seems, cancer.
Several therapeutic avenues for targeting the cellular and molecular components of innate antitumor immunity are being explored. The stimulator of interferon genes (STING) pathway is generating a particular buzz because it is a main promoter of the type I interferon (IFN) response that is central to innate immunity and often defective in cancer cells. Most excitingly, it bridges the 2 arms of the immune system by priming T cells. Therapeutically targeting the STING pathway could transform an immunologically “cold” tumor into a “hot” one, making it more likely to respond to other forms of immunotherapy, such as immune checkpoint inhibitors.
Many pharmaceutical companies are pursuing STING agonists. Although most of the agents are in preclinical or discovery stages, the first clinical trials are under way in combination with checkpoint blockade agents.The innate immune response is the frontline of defense against pathogenic invaders, as well as potentially harmful “self” material. These cellular mediators include natural killer cells, macrophages, neutrophils, and dendritic cells that express receptors encoded by genes inherited though the germline, known as pattern recognition receptors (PRRs).
As their name suggests, these receptors recognize unique molecular patterns found in microorganisms, called pathogen-associated molecular patterns, or components of the host cell that are released during cellular damage or death, called danger-associated molecular patterns (DAMPs). The best-known PRRs are tolllike receptors (TLRs) and C-type lectin receptors, but there are many others, and their engagement depends on the nature of the encountered threat.1
One such threat is foreign nucleic acids; in particular, the microbial and viral DNA that is introduced into the host cell during infection. A potent stimulator of the innate immune response, it triggers the production of type I IFNs (IFN-α and IFN-β) and secretion of proinflammatory cytokines via activation of DNA sensors in the cell’s cytosol.
A vast array of putative cytosolic DNA sensors have been proposed, but how they triggered an IFN response was unclear. In 2008, the STING protein was identified as an essential adapter in this process.
Researchers attempted to link STING to many of these DNA sensors upstream. However, it was not until the discovery in 2013 of a novel DNA sensor, cyclic guanosine monophosphate (GMP)— adenosine monophosphate (AMP) synthase (cGAS), that a clear winner emerged, providing an undisputed biological mechanism by which DNA binding translates into an IFN response.
Other DNA sensors with distinct functions are also well established, but their downstream signaling pathways are poorly understood. These include toll-like receptor (TLR) 9, which predominantly detects endosomal hypomethylated DNA, and sensors of other nucleic acids, such as retinoic acid-inducible gene 1 (RIG-I), TLR7, and TLR8, which detect viral RNA.1-3Current understanding of the STING pathway posits that cGAS senses and binds to cytosolic DNA, triggering a conformational change that activates it. Activated cGAS catalyzes the synthesis of cyclic GMP-AMP (cGAMP) from adenosine triphosphate and guanosine triphosphate. A cyclic dinucleotide (CDN), cGAMP functions as a second messenger, binding to and activating STING, which is found on the membrane of the endoplasmic reticulum.
The precise mechanism through which cGAMP activates STING remains somewhat unclear. Current models suggest that STING proteins are found in pairs and that cGAMP binds to a central crevice in the STING dimer, forcing a structural change in the protein that releases a previously hidden tail.
The STING protein moves to the Golgi apparatus and, through its tail, binds to tank-binding kinase 1 (TBK1). TBK1 phosphorylates transcription factors including IRF3 and NFκB, which move into the nucleus and induce the transcription of type I IFN genes (Figure 1).
STING is found in both immune and nonimmune cells. Within the immune system, it is found on cells of both the innate and adaptive arms, including macrophages, antigen-presenting cells, and T cells. On nonimmune cells, it is found on endothelial cells, fibroblasts, and epithelial cells.
In the context of antigen-presenting cells, the type I IFN response that results from STING activation induces the production of chemokines that attract immune cells. STING also promotes antigen processing and presentation and the priming of CD8-positive T cells. Thus, as the major regulator of the type I IFN response, the STING pathway serves as an important link between the innate and adaptive arms of the immune system (Figure 2).4,5,6More recently, the STING pathway has been shown to be activated by self-DNA in the cytosol, implicating it in the generation of antitumor immunity. Chromosomal and DNA damage is a hallmark of cancer cells, resulting from chromosomal or genome instability inherent in many tumor types or from chemotherapy or radiation therapy.
During cell division, this damaged DNA is often mis-segregated away from the main chromosomal mass and ends up in micronuclei within the cytoplasm. Eventually, the membranes surrounding the micronuclei rupture, spilling the contents into the cytosol, where the nucleic acids act as DAMPs, triggering the STING pathway. Thus, the presence of cytosolic DNA in cancer cells, resulting from genomic damage, initiates an antitumor immune response via the STING pathway.
Unsurprisingly, tumors have evolved countermeasures to downregulate the STING pathway; allow cancer cells to evade this DNA detection pathway; and, ultimately, promote malignancy. A number of mechanisms by which the STING pathway is rendered ineffective have been observed, including reduced expression of the STING and cGAS pathway genes and proteins, as well as epigenetic silencing. What’s more, oncogenic viruses have evolved specific viral proteins that directly antagonize the STING pathway. Both the E7 protein expressed by human papillomavirus and the E1A protein expressed by adenoviruses potently inhibit STING signaling.6,7The STING pathway’s important role in antitumor immunity has spurred efforts to harness the pathway for therapeutic purposes, but this has proved far from straightforward. The first STINGtargeting drug to enter clinical testing was the chemotherapeutic drug 5,6-dimethylxanthenone- 4-acetic acid (DMXAA), which was already being tested in an anticancer capacity prior to the discovery of its STING agonist effects.
Despite potent antitumor effects in mouse models and early-phase clinical trials in patients with non—small cell lung cancer, the drug failed in phase III studies. This was attributed to the smaller sample sizes that overestimated the treatment effect in phase II trials, along with the subsequent finding that DMXAA binds just to the mouse STING protein and does not affect human STING.8
Efforts to target the pathway shifted to the development of STING agonists based on the naturally occurring CDNs that activate human STING. These synthetic CDNs are engineered to be more stable, potent, and effective against the multiple variants of the human STING protein.9
Buoyed by early promise in preclinical trials, a host of pharmaceutical companies have thrown their hats into the ring and are testing their own STING agonists in preclinical trials or have STINGtargeting drugs in the discovery stages. To date, 2 STING agonists have advanced into clinical trials: Aduro Biotech’s ADU-S100 and Merck’s MK-1454, which are being evaluated in phase I studies (Table).10,11
These trials recently began to recruit patients, so the clinical utility of STING agonists remains to be seen. However, preclinical data presented at the 2018 American Association for Cancer Research Annual Meeting highlighted the potent antitumor activity of both of these drugs. GSK532, a STING agonist that GSK is evaluating in the preclinical setting, also showed promising signs of efficacy.
At the highest-tolerated doses, MK-1454 induced complete responses in 100% of mouse tumors. ADU-S100 and GSK532 demonstrated activation of the innate and adaptive immune response and the ability to eliminate or reduce the size of not just the injected tumor but also distal tumors.12-15
Among the key challenges facing STING agonists is the potential for severe toxicity resulting from inducing an IFN response. That, in addition to the short half-life of these drugs, has necessitated intratumoral administration. Another challenge is the poor membrane permeability of CDN-based STING agonists. To be effective, they need to cross the plasma membrane to engage STING in the cytosol. A number of companies are working on alternative methods of administering STING agonists, in addition to exploring liposomal encapsulation to improve delivery across the membrane.16,17
Meanwhile, other companies are focusing on developing STING agonists that are not based on CDNs. For example, both iTeos Therapeutics and Nimbus Therapeutics have a STING program developing small-molecule agonists that have improved drug properties.10,11Investigators also are exploring other components of the nucleic acid—sensing machinery as drug targets. RIG-I, for example, plays a major role in sensing viral RNA and stimulates an IFN response like STING. The RIG-I agonist RGT-100 is in phase I/II clinical trials, as are numerous drugs targeting TLR7, 8, and 9, a subset of TLRs that are involved in sensing pathogenic DNA and RNA.16
Additionally, drugs targeting the killer immunoglobulin receptors, one of the central receptor families involved in the activation of natural killer cells, are progressing through clinical trials. The CD47 protein, dubbed the “don’t eat me” signal because it inhibits the phagocytic activity of macrophages, is also being targeted in this rapidly expanding field.17,18The most excitement for STING agonists surrounds their potential for enhancing the effects of other types of immunotherapy, particularly immune checkpoint inhibitors. These blockbuster drugs exemplify the astounding promise of harnessing the adaptive immune response for cancer immunotherapy and yet have proved ineffective in many patients, partly because they work by reactivating the T cells that have infiltrated the tumor microenvironment, known as tumor-infiltrating lymphocytes (TILs). However, in many types of cancer, the number of TILs is negligible; these are often referred to as non—T-cellinflamed, or immunologically cold, tumors.
Activating the STING pathway not only stimulates the innate antitumor immune response but also acts as a bridge to adaptive immunity by recruiting and priming T cells in the tumor microenvironment. Researchers have suggested, therefore, that it could turn those cold tumors into T-cell-inflamed, immunologically hot tumors that might better respond to immune checkpoint inhibitors and other immunotherapies.10
ADU-S100 and MK-1454 are being investigated in combination with immune checkpoint inhibitors in phase I trials. Preclinical data suggest synergy of these combinations, even in tumors that are intrinsically resistant to immune checkpoint inhibitor monotherapy.12,13
Because of the pathway’s ability to stimulate both an innate and adaptive immune response, STING agonists are also being explored as vaccine adjuvants. In 2015, a cell-based vaccine called STINGVAX was developed, which combines synthetic CDNs with granulocyte-macrophage colony-stimulating factor—producing cells. STINGVAX has shown potent antitumor activity in preclinical models of numerous cancer types, as well as synergy in combination with immune checkpoint inhibitors.19
In another vein, activity of the STING pathway may be a useful biomarker for selecting patients with prostate cancer for immune checkpoint blockade therapies, a malignancy in which this strategy has had limited success.
In research findings presented at the 2018 American Society of Clinical Oncology Annual Meeting, investigators reported on their analyses of nearly 500 prostate cancer tumor samples using the immune-based DNA Damage Repair Deficiency assay, which measures activation of the cGAS/STING pathway.20 When the assay was combined with a test that predicts the risk of metastatic recurrence (the prostate cancer metastatic signature PCM), investigators were able to identify a subset of patients with elevated immune signaling and greater genomic instability. The researchers theorize that this group, which represents 10% to 20% of patients with early prostate cancer, could benefit from immune checkpoint and DNA-damaging therapy.Despite STING’s clear role in mediating antitumor immunity and the preclinical promise of STING agonists, targeting the pathway warrants caution. A deeper understanding of its dynamics has revealed a potentially challenging dichotomy.21,22
Inflammation is a double-edged sword in cancer development. Although clearly an important component of an antitumor immune response, inflammation also undoubtedly is linked to the development and progression of cancer. Thus, it is somewhat unsurprising to find that the STING pathway may also have both a pro- and antitumoral role. These differing effects of STING activation appear to depend on the tumor type and stage.
The STING pathway has also been shown to suppress the antitumor immune response via a number of mechanisms, including induction of the expression of indoleamine 2,3-dioxygenase, which activates regulatory T cells.23
Most recently, the STING pathway has been implicated in the promotion of metastasis. The spread of cancers is responsible for more than 90% of cancer-related deaths24; thus, understanding how it occurs is among the central challenges to effective anticancer treatment.
A recent study published in Nature identified chromosomal instability as a widespread feature of metastatic cancer cells. Cells with high levels of chromosomal instability were more likely to metastasize than those with lower levels. While searching for an explanation for this phenomenon, researchers found that metastatic cells also had increased expression of genes that controlled inflammatory responses.
Further investigation revealed that chromosomal instability caused DNA to leak into the cytosol and triggered the STING pathway, leading to a chronic inflammatory response. In normal cells, this should have led to cell death; however, the study authors suggested that the cancer cells were mimicking immune cells such as macrophages and responding to the inflammatory response by becoming migratory, thereby promoting their metastatic capabilities.21,25